Liwei Lin
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Research Interests
Recognitions
Liwei Lin has a total of 555 co-authors. Here are the top co-authors by number of shared publications:
Mohan Sanghadasa
Aviation and Missile Center, U.S. Army Combat Capabilities Development Command, Redstone Arsenal, AL, 35898, USA
13 shared publications- Xining Zang
State Key Laboratory of Clean and Efficient Turbomachinery Power Equipment, Department of Mechanical Engineering, Tsinghua University, Beijing, China
13 shared publications - Caiwei Shen
Mechanical Engineering Department, University of Massachusetts Dartmouth, Dartmouth, MA, USA
11 shared publications - Junwen Zhong
Department of Electromechanical Engineering and Centre for Artificial Intelligence and Robotics, University of Macau, Macau SAR, China
8 shared publications Xiaohao Wang
Tsinghua Shenzhen International School, Tsinghua University, Shenzhen, 518055, China
6 shared publications- Yande Peng
Department of Mechanical Engineering, University of California, Berkeley, USA
6 shared publications
Latest Publications
- Top Cited PapersOpen Access
- 2 April 2026
- Published by Springer Nature in Nature Communications
- Vol. 17 (1)
Semi-transparent and eco-friendly face masks for PM0.3 filtration via biomimetic-fractal nanofibers
To address the need for sustainable high-performance face masks, we demonstrate a strategy that utilizes zein for efficient nanofabrication. An electrospinning process controllably splits the fluid jet, allowing for the creation of face masks composed of fine, fractal-like nanofibers without the need for impurity excipients. This process, enabled by an dipole-force-induced interfacial fluctuation mechanism that we identified, results in face masks that achieve exceptional air filtration performance, surpassing that of commercial N95 masks, while reducing both weight and thickness by half. The significant reduction in material use grants the face masks high optical transparency, facilitating unobstructed facial recognition. Crucially, the pure zein composition allows for the reprocessing of the face masks, contributing to further resource conservation. This work provides a simple, eco-friendly solution for advanced personal protection and points the way toward sustainable wearable devices.
- 2 April 2026
- Published by Springer Nature in Nature Communications
- Vol. 17 (1)
Top Cited PapersOpen AccessSemi-transparent and eco-friendly face masks for PM0.3 filtration via biomimetic-fractal nanofibers
To address the need for sustainable high-performance face masks, we demonstrate a strategy that utilizes zein for efficient nanofabrication. An electrospinning process controllably splits the fluid jet, allowing for the creation of face masks composed of fine, fractal-like nanofibers without the need for impurity excipients. This process, enabled by an dipole-force-induced interfacial fluctuation mechanism that we identified, results in face masks that achieve exceptional air filtration performance, surpassing that of commercial N95 masks, while reducing both weight and thickness by half. The significant reduction in material use grants the face masks high optical transparency, facilitating unobstructed facial recognition. Crucially, the pure zein composition allows for the reprocessing of the face masks, contributing to further resource conservation. This work provides a simple, eco-friendly solution for advanced personal protection and points the way toward sustainable wearable devices.
- Open Access
AI-Assisted Hyperspectral Interferometry and Single-Cell Dispersion Imaging
Interferometry techniques are essential for extracting phase information from optical systems, enabling precise measurements of dispersion and highly sensitive detection of perturbations. While phase sensing offers enhanced sensitivity compared to conventional spectroscopy methods, this sensitivity often makes systems more vulnerable to external factors such as vibrations, introducing instability and noise. In this work, we demonstrate a broadband and AI-enhanced interferometry method, denoted general polarization common-path interferometry (GPCPI), that relaxes the polarization constraints of traditional common-path interferometry. The polarization decoupling feature enables simultaneous amplitude and phase measurements supplemented with deep neural autoencoders to detect phase anomalies in the spectrum through the analysis of second order derivative mapping of the phase profile, enhancing the accuracy of broadband phase measurements. The approach enables an order of magnitude improvement in phase stability compared to state-of-the-art interferometry techniques, leading to higher accuracy in phase sensing. Plasmonic metasurface phase sensing and hyperspectral single-cell dispersion imaging demonstrate the capability and sensitivity of the method over conventional spectroscopy. Our own adopted version of deep learning model, ConvNeXt V2, enables single-shot and real-time tracking of phase variation with minimized noise. Interference fringes over the cell cultured samples reveal the fingerprints of the normal (CCD-32Sk) vs cancerous (COLO-829) skin cells, enabling robust cell classification and disease diagnosis at single-cell level. The proposed interferometry technique offers a reliable, compact, and stable solution for broadband phase measurements and single-cell dispersion imaging for applications in metrology, molecular diagnostics, drug discovery, and quantum sensing.
Open AccessAI-Assisted Hyperspectral Interferometry and Single-Cell Dispersion Imaging
Interferometry techniques are essential for extracting phase information from optical systems, enabling precise measurements of dispersion and highly sensitive detection of perturbations. While phase sensing offers enhanced sensitivity compared to conventional spectroscopy methods, this sensitivity often makes systems more vulnerable to external factors such as vibrations, introducing instability and noise. In this work, we demonstrate a broadband and AI-enhanced interferometry method, denoted general polarization common-path interferometry (GPCPI), that relaxes the polarization constraints of traditional common-path interferometry. The polarization decoupling feature enables simultaneous amplitude and phase measurements supplemented with deep neural autoencoders to detect phase anomalies in the spectrum through the analysis of second order derivative mapping of the phase profile, enhancing the accuracy of broadband phase measurements. The approach enables an order of magnitude improvement in phase stability compared to state-of-the-art interferometry techniques, leading to higher accuracy in phase sensing. Plasmonic metasurface phase sensing and hyperspectral single-cell dispersion imaging demonstrate the capability and sensitivity of the method over conventional spectroscopy. Our own adopted version of deep learning model, ConvNeXt V2, enables single-shot and real-time tracking of phase variation with minimized noise. Interference fringes over the cell cultured samples reveal the fingerprints of the normal (CCD-32Sk) vs cancerous (COLO-829) skin cells, enabling robust cell classification and disease diagnosis at single-cell level. The proposed interferometry technique offers a reliable, compact, and stable solution for broadband phase measurements and single-cell dispersion imaging for applications in metrology, molecular diagnostics, drug discovery, and quantum sensing.
- Open Access
- 5 December 2025
- Published by Wiley in Advanced Materials Technologies
- Vol. 11 (13)
Advanced Design for Weakly Coupled Resonators by Automatic Active Optimization
Advanced design for weakly coupled resonators (WCRs) based on micro disk‐shaped structures has been developed using an automatic active optimization (AAO) process with low coupling factors. Previous design methodologies have often used supervised learning (SL) with genetic algorithm for the optimization process but the limited initial design domain in the SL‐based predictor often results in local instead of global optimizations. This AAO scheme can generate superior designs in the neighborhood of initial high‐density dataset regions and add newly‐found designs to the original dataset to shift the data distribution toward a more generalized model to seek global optimal designs. Results show that after 10 to 15 epochs, the accuracy of the predictors for the first‐order mode and the second‐order mode coupling factor enhances significantly, with the root‐mean‐square errors dropping by 84.9% and 83.8%, respectively. Simultaneously, the optimized designs have resulted in reduced coupling factors of 0.033 and 5.7 × 10 −11 for the first‐ and second‐order modes, respectively, as compared to those in the initial datasets.
- 5 December 2025
- Published by Wiley in Advanced Materials Technologies
- Vol. 11 (13)
Open AccessAdvanced Design for Weakly Coupled Resonators by Automatic Active Optimization
Advanced design for weakly coupled resonators (WCRs) based on micro disk‐shaped structures has been developed using an automatic active optimization (AAO) process with low coupling factors. Previous design methodologies have often used supervised learning (SL) with genetic algorithm for the optimization process but the limited initial design domain in the SL‐based predictor often results in local instead of global optimizations. This AAO scheme can generate superior designs in the neighborhood of initial high‐density dataset regions and add newly‐found designs to the original dataset to shift the data distribution toward a more generalized model to seek global optimal designs. Results show that after 10 to 15 epochs, the accuracy of the predictors for the first‐order mode and the second‐order mode coupling factor enhances significantly, with the root‐mean‐square errors dropping by 84.9% and 83.8%, respectively. Simultaneously, the optimized designs have resulted in reduced coupling factors of 0.033 and 5.7 × 10 −11 for the first‐ and second‐order modes, respectively, as compared to those in the initial datasets.
- 7 November 2025
- Published by Springer Nature in Nature Electronics
- Vol. 8 (11) , 1059-1071
Laser-assisted direct three-dimensional printing of free-standing thermoset devices
- 7 November 2025
- Published by Springer Nature in Nature Electronics
- Vol. 8 (11) , 1059-1071
Laser-assisted direct three-dimensional printing of free-standing thermoset devices
- Open Access
- 6 November 2025
- Published by Springer Nature in Microsystems & Nanoengineering
- Vol. 11 (1) , 1-12
Subcutaneous and continuous blood pressure monitoring in an ambulatory sheep by piezoelectric micromachined ultrasonic transducers
This paper presents subcutaneous and continuous blood pressure (BP) monitoring using aluminum nitride (AlN) piezoelectric micromachined ultrasonic transducers (PMUTs) in an ambulatory sheep. A 37 $$\times$$ 45 PMUTs array with a footprint of 5 $$\times$$ 5 mm 2 has been designed and fabricated as a prototype device. The deep reactive ion etching (DRIE) process to open the backside holes on the silicon substrate has been optimized to create active device diaphragms with a radius of 29 μm. The resulting PMUT unit has a measured resonant frequency of 6.5 MHz in water, an output acoustic pressure of 28 kPa at a distance of 10 mm, and a 6-dB bandwidth of about 33%. The BP monitoring scheme is validated through both in vitro and in vivo experiments to illustrate the correlation between the diameter of the blood vessel and pressure. Simulations indicate that possible issues in misalignment between the device and the blood vessel can result in a 60% reduction in signal strength with only 1 mm in misalignment. This highlights the advantage of subcutaneous implantation in maintaining a stable interface and consistent alignment for reliable long-term BP monitoring, in contrast to similar approaches via wearable system setups. The in vivo testing result shows BP wave fine features such as dicrotic notches and the averaged systolic/diastolic pressure errors are −1.2 $$\pm$$ 2.1 and −2.9 $$\pm$$ 1.4 mmHg, respectively, which meets the clinical standard as calibrated by a gold-standard arterial line pressure sensor. As such, this system highlights the potential applications in silent, continuous, and highly accurate BP monitoring for hypertension patients using this implantable MEMS-based technology.- 6 November 2025
- Published by Springer Nature in Microsystems & Nanoengineering
- Vol. 11 (1) , 1-12
Open AccessSubcutaneous and continuous blood pressure monitoring in an ambulatory sheep by piezoelectric micromachined ultrasonic transducers
This paper presents subcutaneous and continuous blood pressure (BP) monitoring using aluminum nitride (AlN) piezoelectric micromachined ultrasonic transducers (PMUTs) in an ambulatory sheep. A 37 $$\times$$ 45 PMUTs array with a footprint of 5 $$\times$$ 5 mm 2 has been designed and fabricated as a prototype device. The deep reactive ion etching (DRIE) process to open the backside holes on the silicon substrate has been optimized to create active device diaphragms with a radius of 29 μm. The resulting PMUT unit has a measured resonant frequency of 6.5 MHz in water, an output acoustic pressure of 28 kPa at a distance of 10 mm, and a 6-dB bandwidth of about 33%. The BP monitoring scheme is validated through both in vitro and in vivo experiments to illustrate the correlation between the diameter of the blood vessel and pressure. Simulations indicate that possible issues in misalignment between the device and the blood vessel can result in a 60% reduction in signal strength with only 1 mm in misalignment. This highlights the advantage of subcutaneous implantation in maintaining a stable interface and consistent alignment for reliable long-term BP monitoring, in contrast to similar approaches via wearable system setups. The in vivo testing result shows BP wave fine features such as dicrotic notches and the averaged systolic/diastolic pressure errors are −1.2 $$\pm$$ 2.1 and −2.9 $$\pm$$ 1.4 mmHg, respectively, which meets the clinical standard as calibrated by a gold-standard arterial line pressure sensor. As such, this system highlights the potential applications in silent, continuous, and highly accurate BP monitoring for hypertension patients using this implantable MEMS-based technology. - Open Access
- 27 September 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-21
MorphingSkin: A Skin-like Platform that Integrates Multimodal Hydraulic Actuators Based on Flexible Electroosmotic Pumps
- 27 September 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-21
Open AccessMorphingSkin: A Skin-like Platform that Integrates Multimodal Hydraulic Actuators Based on Flexible Electroosmotic Pumps
- Open Access
- 22 August 2025
- Published by American Association for the Advancement of Science (AAAS) in Science Advances
- Vol. 11 (34) , eady5008
Hypersensitive pressure sensors inspired by scorpion mechanosensory mechanisms for near-body flow detection in intelligent robots
Sensitivity enhancement for pressure sensors over a broad linear range can improve sensing performance for a wide range of applications such as health monitoring and artificial intelligence. Here, inspired by the high-precision mechanosensory mechanism of the scorpion, a bioinspired piezoresistive pressure sensor (BPPS) is reported for the synergistic enhancement of sensitivity and linearity at 65.56 millivolts per volt per kilopascal and 0.99934, respectively, in a pressure range from 0 to 500 kilopascals. The BPPS can distinguish laminar, transitional, and turbulent flows as well as identify approaching objects of different shapes with an accuracy exceeding 85.42% by integrating a wavelet transform algorithm and the ResNet18 deep learning network. As a proof of concept, BPPS has been engineered in a hexapod robot to enable near-body flow field sensing for active collision avoidance. This work underscores the potential to leverage key design concepts inspired by living insects for improved sensing performance and offers structural insights for other high-precision sensors.
- 22 August 2025
- Published by American Association for the Advancement of Science (AAAS) in Science Advances
- Vol. 11 (34) , eady5008
Open AccessHypersensitive pressure sensors inspired by scorpion mechanosensory mechanisms for near-body flow detection in intelligent robots
Sensitivity enhancement for pressure sensors over a broad linear range can improve sensing performance for a wide range of applications such as health monitoring and artificial intelligence. Here, inspired by the high-precision mechanosensory mechanism of the scorpion, a bioinspired piezoresistive pressure sensor (BPPS) is reported for the synergistic enhancement of sensitivity and linearity at 65.56 millivolts per volt per kilopascal and 0.99934, respectively, in a pressure range from 0 to 500 kilopascals. The BPPS can distinguish laminar, transitional, and turbulent flows as well as identify approaching objects of different shapes with an accuracy exceeding 85.42% by integrating a wavelet transform algorithm and the ResNet18 deep learning network. As a proof of concept, BPPS has been engineered in a hexapod robot to enable near-body flow field sensing for active collision avoidance. This work underscores the potential to leverage key design concepts inspired by living insects for improved sensing performance and offers structural insights for other high-precision sensors.
- Open Access
- 11 August 2025
- Published by Springer Nature in npj Computational Materials
- Vol. 11 (1) , 1-10
Multi-target digital material design via a conditional denoising diffusion probability model
Multi-target digital material design has been challenging due to the expansive design space and instability of traditional methods in satisfying multiple objectives. This work proposes and demonstrates a customizer based on a classifier-free, conditional denoising diffusion probability model (cDDPM) to efficiently create the layouts of digital materials meeting the design goal of multiple mechanical properties all together. A case study has been conducted based on a micro mechanical resonator with four pre-assigned resonant frequencies. Using 29,430 samples generated via finite element analysis (FEA), the cDDPM is trained to simultaneously customize up to four vibrational modes, achieving over 95% prediction accuracy. Furthermore, the cDDPM approach also shows superior performances in the single-target customization for up to 99% in prediction accuracy when compared with traditional conditional generative adversarial networks (cGANs). As such, the proposed design framework provides a highly customizable and robust methodology for the design of complicated digital materials.
- 11 August 2025
- Published by Springer Nature in npj Computational Materials
- Vol. 11 (1) , 1-10
Open AccessMulti-target digital material design via a conditional denoising diffusion probability model
Multi-target digital material design has been challenging due to the expansive design space and instability of traditional methods in satisfying multiple objectives. This work proposes and demonstrates a customizer based on a classifier-free, conditional denoising diffusion probability model (cDDPM) to efficiently create the layouts of digital materials meeting the design goal of multiple mechanical properties all together. A case study has been conducted based on a micro mechanical resonator with four pre-assigned resonant frequencies. Using 29,430 samples generated via finite element analysis (FEA), the cDDPM is trained to simultaneously customize up to four vibrational modes, achieving over 95% prediction accuracy. Furthermore, the cDDPM approach also shows superior performances in the single-target customization for up to 99% in prediction accuracy when compared with traditional conditional generative adversarial networks (cGANs). As such, the proposed design framework provides a highly customizable and robust methodology for the design of complicated digital materials.
- 14 July 2025
- Published by Wiley in Advanced Materials
- Vol. 37 (39) , e2507859
Highly Responsive Self‐Healing and Degradable Piezoelectric Soft Machines
Piezoelectric materials that are simultaneously healable, stretchable, and degradable have remained an unmet challenge, limiting advancements in wearable and implantable electronics, where devices face multidimensional mechanical deformation, causing a risk of damage. To address this critical gap, a biocompatible piezoelectric material is developed for ultrahigh piezoelectric effects with DL‐alanine amino acid crystals, which is stretchable, healable, and degradable. The in situ grown DL‐alanine piezoelectric crystals within an ionically cross‐linked gelatin hydrogel matrix strengthen the piezoelectric properties with an ultrahigh voltage coefficient of 1.6 Vm N −1 . The combination of the piezo‐ionic property and crystal alignment results in a record‐breaking energy harvesting figure‐of‐merit value at 57.6 pm 2 N −1 to deliver outstanding mili‐watt level power outputs in proof‐of‐concept devices which can power up even several electric light bulbs. An elastically stretchable, damage resistant strain sensor is further optimized for real‐time healthcare monitoring and biomechanical motion tracking. By integrating machine learning algorithms, the sensing system intelligently classifies biomechanical activities with high accuracy, enabling advanced applications in healthcare, rehabilitation, and sports monitoring.
- 14 July 2025
- Published by Wiley in Advanced Materials
- Vol. 37 (39) , e2507859
Highly Responsive Self‐Healing and Degradable Piezoelectric Soft Machines
Piezoelectric materials that are simultaneously healable, stretchable, and degradable have remained an unmet challenge, limiting advancements in wearable and implantable electronics, where devices face multidimensional mechanical deformation, causing a risk of damage. To address this critical gap, a biocompatible piezoelectric material is developed for ultrahigh piezoelectric effects with DL‐alanine amino acid crystals, which is stretchable, healable, and degradable. The in situ grown DL‐alanine piezoelectric crystals within an ionically cross‐linked gelatin hydrogel matrix strengthen the piezoelectric properties with an ultrahigh voltage coefficient of 1.6 Vm N −1 . The combination of the piezo‐ionic property and crystal alignment results in a record‐breaking energy harvesting figure‐of‐merit value at 57.6 pm 2 N −1 to deliver outstanding mili‐watt level power outputs in proof‐of‐concept devices which can power up even several electric light bulbs. An elastically stretchable, damage resistant strain sensor is further optimized for real‐time healthcare monitoring and biomechanical motion tracking. By integrating machine learning algorithms, the sensing system intelligently classifies biomechanical activities with high accuracy, enabling advanced applications in healthcare, rehabilitation, and sports monitoring.
- Open Access
- 17 May 2025
- Published by Springer Nature in Nature Communications
- Vol. 16 (1) , 1-13
Plasmonic coffee-ring biosensing for AI-assisted point-of-care diagnostics
A major challenge in addressing global health issues is developing simple, affordable biosensors with high sensitivity and specificity. Significant progress has been made in at-home medical detection kits, especially during the COVID-19 pandemic. Here, we demonstrated a coffee-ring biosensor with ultrahigh sensitivity, utilizing the evaporation of two sessile droplets and the formation of coffee-rings with asymmetric nanoplasmonic patterns to detect disease-relevant proteins as low as 3 pg/ml, under 12 min. Experimentally, a protein-laden droplet dries on a nanofibrous membrane, pre-concentrating biomarkers at the coffee ring. A second plasmonic droplet with functionalized gold nanoshells is then deposited at an overlapping spot and dried, forming a visible asymmetric plasmonic pattern due to distinct aggregation mechanisms. To enhance detection sensitivity, a deep neural model integrating generative and convolutional networks was used to enable quantitative biomarker diagnosis from smartphone photos. We tested four different proteins, Procalcitonin (PCT) for sepsis, SARS-CoV-2 Nucleocapsid (N) protein for COVID-19, Carcinoembryonic antigen (CEA) and Prostate-specific antigen (PSA) for cancer diagnosis, showing a working concentration range over five orders of magnitude. Sensitivities surpass equivalent lateral flow immunoassays by over two orders of magnitude using human saliva samples. The detection principle, along with the device, and materials can be further advanced for early disease diagnostics.
- 17 May 2025
- Published by Springer Nature in Nature Communications
- Vol. 16 (1) , 1-13
Open AccessPlasmonic coffee-ring biosensing for AI-assisted point-of-care diagnostics
A major challenge in addressing global health issues is developing simple, affordable biosensors with high sensitivity and specificity. Significant progress has been made in at-home medical detection kits, especially during the COVID-19 pandemic. Here, we demonstrated a coffee-ring biosensor with ultrahigh sensitivity, utilizing the evaporation of two sessile droplets and the formation of coffee-rings with asymmetric nanoplasmonic patterns to detect disease-relevant proteins as low as 3 pg/ml, under 12 min. Experimentally, a protein-laden droplet dries on a nanofibrous membrane, pre-concentrating biomarkers at the coffee ring. A second plasmonic droplet with functionalized gold nanoshells is then deposited at an overlapping spot and dried, forming a visible asymmetric plasmonic pattern due to distinct aggregation mechanisms. To enhance detection sensitivity, a deep neural model integrating generative and convolutional networks was used to enable quantitative biomarker diagnosis from smartphone photos. We tested four different proteins, Procalcitonin (PCT) for sepsis, SARS-CoV-2 Nucleocapsid (N) protein for COVID-19, Carcinoembryonic antigen (CEA) and Prostate-specific antigen (PSA) for cancer diagnosis, showing a working concentration range over five orders of magnitude. Sensitivities surpass equivalent lateral flow immunoassays by over two orders of magnitude using human saliva samples. The detection principle, along with the device, and materials can be further advanced for early disease diagnostics.